A building fire leaves a physical record of how it grew, even after it has burned through several rooms. Reading that record correctly is the entire discipline.
Start a conversation with our AI Research Concierge, already scoped to structure fire. Pick a starting point, or describe your situation directly.
Structure fire investigation is the origin-and-cause discipline in its most general form — determining where a fire started and why, independent of whether the eventual cause turns out to be electrical, mechanical, or human. The physical evidence is fire pattern: the char depth, calcination, and burn severity distributed across a room record the fire's growth and direction, but only if they are read correctly. A room that reaches flashover, the point at which everything in it ignites nearly simultaneously, produces post-flashover damage that can mimic multiple origins even though the fire started in one place. Getting the origin right, and then correctly separating an accidental cause from an incendiary one without relying on the discredited practice of concluding arson simply because no accidental cause could be found, is where this work is won or lost.
Origin determination and fire behavior are inseparable — how a fire grew shapes what evidence survives to show where it started.
Correlating fire patterns, char depth, and burn severity across the scene to isolate the specific area where the fire began.
Ventilation-limited versus fuel-limited growth, and how post-flashover conditions in a fully involved room can mask or mimic multiple points of origin.
V-patterns, low burn patterns, and hourglass patterns, interpreted against their actual governing physics rather than the older rules of thumb that produced them.
Distinguishing genuine accelerant evidence from alternate innocent explanations, and rejecting "negative corpus" — concluding arson solely because no accidental cause was found — as invalid under current NFPA 921.
Sprinkler, detection, and compartmentation performance, or failure, as a driver of how far the fire spread and how large the resulting loss became.
Zone or computational fire models testing whether a proposed origin and growth scenario is physically consistent with the observed damage and any witness timeline.
Structure fire investigation moves systematically from the exterior inward and from the least to the most damaged areas, so early conclusions do not bias what gets examined next.
A structure fire opinion carries consequences well beyond the property loss itself:
Debris that looks like trash to a cleanup crew can be the origin-area evidence that determines cause. Secure the structure and hold demolition until the origin area has been documented and, where relevant, sampled.
Investigators map fire patterns — char depth, calcination, and burn severity — across every affected room and use the relative severity gradient to work backward toward the area with the earliest and most severe damage. The building's ventilation openings, structural collapse, and any witness or 911-call timeline are weighed alongside the pattern evidence, since a fire that reached flashover can spread damage in ways that are not simply radial from the origin.
Negative corpus was the practice of concluding a fire was incendiary because investigators could not identify an accidental cause, reasoning that the absence of an accidental explanation proved arson. The 2011 edition of NFPA 921 explicitly rejected this reasoning because it inverts the scientific method — a hypothesis has to be affirmatively supported by evidence, not accepted by default because its alternatives were eliminated. An opinion built on negative corpus reasoning is now a well-recognized vulnerability in litigation.
No single pattern is conclusive on its own, and several patterns once treated as reliable incendiary indicators — including certain glass crazing and some low-burn signatures — have since been shown by fire testing to occur under ordinary accidental conditions as well. An incendiary determination requires positive evidence: a confirmed ignitable liquid, multiple unconnected origins, or physical evidence of a deliberate ignition device, evaluated together rather than any single pattern in isolation.
Flashover is the point at which radiant heat in a compartment ignites all exposed combustible surfaces nearly simultaneously, converting a localized fire into full room involvement within seconds. After flashover, the entire room burns with roughly equal severity, which can erase the pattern gradient that would otherwise point back to where the fire started, and can create secondary burn patterns that look like independent origins even though they are a downstream effect of the same event.
Trained accelerant-detection canines are a useful and widely used field screening tool, with a strong record of directing sample collection toward areas worth testing. But a canine alert is not, by itself, confirmatory evidence — laboratory GC-MS analysis of the actual debris sample is what determines whether an ignitable liquid is present and identifies what it is. Relying on a canine alert without laboratory confirmation is a recognized weak point in an incendiary determination.
Technical briefings from our work in this area.
A canine alert is a screening tool that directs sampling. It is not a laboratory result, and the difference between the two is a recurring point of failure in fire investigations.
readConcluding a cause by eliminating every alternative is not the same as proving it. The distinction between legitimate elimination and reasoning from ignorance is where many cause opinions are lost.
readFlashover degrades the fire patterns that origin determination depends on. What survives, what does not, and how a defensible origin opinion is built when the compartment burned to completion.
readTell us what occurred. We will triage it and connect you with the right expert — usually within one business day.